Carbohydrates can be used as vaccines if they are covalently conjugated to a carrier protein. However, some carbohydrate antigens remain poor immunogens even when coupled to a protein. To overcome this weakness, chemical alteration of the native carbohydrate has been proposed. While these modifications can lead to enhanced immunogenicity, the repertoire of antibodies may have a compromised ability to recognize the original native antigen. Consistent native-like immune responses depend on maintaining the same 3D epitope in the vaccine as that in the native antigen. Experimental methods such as NMR spectroscopy and protein crystallography are the cornerstones of 3D structural characterization, but each method faces difficulties when applied to highly flexible carbohydrates and polysaccharides and their complexes with proteins. Here we will use computational methods to provide models for these complexes, using a combination of automated docking, molecular dynamics (MD) and thermodynamic integration (TI) simulations, and test the validity of these models by comparison with existing 3D structural data for carbohydrate-antibody complexes. Concurrently, we will apply the emerging experimental technique of oxidative protein surface footprinting [1, 2], to generate high-throughput, medium resolution experimental data for use as structural constraints to guide the computational docking. These studies will provide a comprehensive, validated, relatively rapid and high-throughput approach characterizing the 3D structures of antibody carbohydrate complexes, which will additionally enable us to characterize the impacts of chemical modifications on the 3D properties of carbohydrate antigens.